舰载直升机在起降和飞行过程中需要面对极为复杂的气动环境，这种环境具有高度的非定常性和不确定性，给直升机的飞行安全带来了巨大的挑战, 但现有的研究手段存在流场数据量大、数值计算成本高等问题。且近年来，深度学习、降阶模型等数据驱动方法在空气动力学等工程应用领域的研究开始逐渐深入。鉴于此，本文首先提出了具有拓展性的多任务学习网络MT-Swin-T，实现了旋翼翼型流场与升力系数的联合预测。其次，分别通过POD-RBF和POD-MLP模型预测旋翼非定常气动力数据，展示了降阶模型在提升计算效率方面的优势。接着还通过POD、SPOD方法对舰艉流场的空间和时频特性进行了降阶分析，提出了基于AE-RBF和PODAE-RBF的舰艉流场预测模型，能够在不同工况下准确预测舰艉流场特征。并基于 POD 和 SPOD 降阶模型，建立了直升机非定常载荷滤波方法，研究保留模态数和不同频率成分对直升机非定常载荷水平的影响。此外，结合降阶预测模型，提出了连续性风限图计算方法，并分析了不同着舰方式对着舰飞行安全边界的影响。主要工作内容包括以下几个方面：

(1) 本文提出了一种多任务深度学习框架MT-Swin-T，结合Swin-Transformer和MLP网络，实现了旋翼翼型的多任务学习和流场预测。MT-Swin-T通过硬参数共享的方式，针对不同任务使用不同的输出结构，即共享Swin-Transformer的编码器部分学习翼型的形状和初始条件的特性，解码器则负责预测速度场信息，同时使用MLP预测升力系数。研究通过实验分析了数据集规模、模型参数、不同损失函数组合等因素对模型精度的影响。实验表明，较大规模的数据集和匹配的模型参数能够显著提升模型的收敛效率与预测精度。相较于传统CFD模拟方法，MT-Swin-T不仅缩短了计算时间，还能在多任务的情况下提供更加准确的预测结果。

(2) 第三章利用POD方法对AH-1G旋翼的非定常气动力数据进行了降阶处理，提取了主要模态，并提出了POD-RBF和POD-MLP两种降阶预测模型。这两种模型分别使用径向基函数方法和多层感知机对测试集中样本的POD模态系数进行预测，并重构出完整的气动力数据。实验表明，POD-RBF模型在载荷波动较大的区域表现出较好的误差控制，而POD-MLP模型在更复杂的飞行工况下误差较大，泛化能力相对较弱。

(3) 针对非定常舰艉流场提出了多种降阶预测模型。首先利用高精度DES方法计算出舰艉非定常流场数据，然后分别采用上一章发展的POD方法和SPOD方法对流场数据进行降阶，对不同风向角条件下艉流场能量特性与频率特性进行了分析。并研究发展了AE-RBF模型，用于时均舰艉流场的降阶预测。此外，还结合POD与自编码器建立了PODAE-RBF模型，用于非定常舰艉流场的降阶预测。结果表明，AE-RBF和PODAE-RBF模型能够在不同风向角和来流风速下准确预测流场特性，且随着潜变量维度的增加，预测效果进一步提升。

(4) 在直升机/舰船动态界面的分析中，本文采用了POD数据降阶传递方法和SPOD非定常载荷滤波方法。POD数据降阶方法通过选取少数主要模态，有效表达舰艉流场的时均特性，大幅降低了流场数据规模，并能够准确模拟直升机着舰进场过程中操纵杆量的变化。SPOD方法则用于分离出不同频率下的模态特征，能够有效捕捉直升机非定常载荷中的低频能量分布。通过保留0.2-3Hz频率成分的SPOD重构流场数据，研究实现了对直升机非定常载荷水平的准确预测。

(5) 本文结合直升机平衡特性的理论风限图计算方法，以及第四章提出的舰艉流场降阶预测模型，发展出了一套新的基于降阶模型的连续性风限图计算方法。此方法通过降阶预测模型ShipROM实时获取时均舰艉流场，极大提高了计算效率，省去了传统CFD方法中多次重新计算舰艉流场的步骤。研究还分析了不同着舰路径对飞行安全边界的影响，结果表明，从甲板后方进场时的飞行安全边界较大，尤其是直升机机头迎风进场时安全性最高，而侧向进场时，尤其是右舷侧进场，尾桨受侧洗速度影响较大，飞行员操纵负荷增加，飞行安全边界相对较小。

Shipboard helicopters encounter extremely complex aerodynamic environments during take-off, landing, and flight, characterized by high levels of unsteadiness and uncertainty, which pose significant challenges to flight safety.However, the existing research methods have some problems, such as large amount of flow field data and high cost of numerical calculation. In recent years, data-driven methods such as deep learning and reduced-order models have gradually gained traction in fields like aerodynamics. In light of this, this thesis first proposes an extensible multi-task learning network, MT-Swin-T, which achieves joint prediction of rotor airfoil flow fields and lift coefficients. Then, POD-RBF and POD-MLP models are employed to predict the unsteady aerodynamic forces of the helicopter rotor, demonstrating the advantages of reduced-order models in improving computational efficiency. Additionally, the spatial and spatio-temporal characteristics of the ship airwake are analyzed with POD and SPOD methods, and AE-RBF and PODAE-RBF-based prediction models are proposed to accurately predict ship airwake characteristics under various innitial conditions. Based on the POD and SPOD reduced-order models, a filtering method for unsteady helicopter loads is developed, examining the influence of retained modes and different frequency components of airwake on unsteady helicopter load levels. Furthermore, a continuous SHOL(ship-helicopter operating limits) calculation method based on reduced-order prediction models is proposed, and the impact of different landing approaches on SHOL is analyzed. The main contributions include the following:

(1) This investigation proposes a multi-task deep learning framework, MT-Swin-T, which combines the Swin-Transformer and MLP networks to achieve multi-task learning and flow field prediction for rotor airfoils. MT-Swin-T employs hard parameter sharing, where the Swin-Transformer encoder learns the airfoil's shape and initial condition features, while different output structures are used for different tasks. The decoder predicts flow velocity fields, and the MLP predicts lift coefficients. Experimental results indicate that larger datasets and matched model parameters significantly enhance convergence efficiency and prediction accuracy. Compared to traditional CFD methods, MT-Swin-T reduces computation time and provides more accurate predictions in multi-task conditions.

(2) This chapter uses POD to process aerodynamic data of the AH-1G rotor, extracting primary modes, and proposes two reduced-reconstruction models: POD-RBF and POD-MLP. These models use radial basis functions and multi-layer perceptrons to predict POD modal coefficients of test samples and reconstruct complete aerodynamic data. The results show that the POD-RBF model exhibits better accuracy in regions with significant load fluctuations, while the POD-MLP model shows higher errors and limited generalization in more complex flight conditions.

(3) For the unsteady ship airwake, multiple reduced-order prediction models were developed. High-fidelity DES simulations were used to simulate the unsteady airwake flow field, followed by POD and SPOD techniques to analyze energy and frequency characteristics under different wind angles. The AE-RBF model was developed for predicting time-averaged airwake flow fields, and a PODAE-RBF model was established for unsteady airwake flow fields. Results indicate that AE-RBF and PODAE-RBF models can accurately predict flow characteristics under varying wind directions and inflow velocities, with improved predictive performance as latent variable dimensions increase.

(4) In the analysis of helicopter/ship dynamic interface, the investigation employs POD data reduction transmission and SPOD unsteady load filtering methods. The POD method captures the time-averaged characteristics of the ship airwake flow field using a small number of primary modes, significantly reducing the data size while accurately simulating helicopter control inputs during landing approaches. The SPOD method extracts modal characteristics across different frequencies and effectively captures low-frequency energy distribution in unsteady helicopter loads. By retaining SPOD frequency components in the 0.2-3 Hz range, the study achieves accurate predictions of unsteady helicopter load levels.

(5) The analysis integrates the theoretical SHOL calculation method for helicopter trimming characteristics with the reduced-order airwake flow field prediction models from Chapter 4, developing a continuous SHOL calculation method based on reduced-order models. This method, ShipROM, obtains time-averaged ship airwake flow fields in real-time, greatly improving computational efficiency by eliminating the need for multiple re-simulations of the airwake flow field in traditional CFD methods. The study also analyzes the impact of different landing paths on SHOL, showing that rear-deck landing approaches have larger safety boundaries, especially when the helicopter nose is facing into the wind. In contrast, sideward approaches, particularly on the starboard side, reduce safety boundaries due to increased pilot workload from tail rotor sidewash effects.

[1] Li K, Kou J, Zhang W. Deep Learning for Multifidelity Aerodynamic Distribution Modeling from Experimental and Simulation Data[J]. AIAA Journal, 2022: 1-15.

[2] Portal-Porras K, Fernandez-Gamiz U, Zulueta E, et al. CNN-based flow control device modelling on aerodynamic airfoils[J]. Scientific Reports, 2022, 12.

[3] White C, Ushizima D, Farhat C. Fast Neural Network Predictions from Constrained Aerodynamics Datasets[C]//AIAA Scitech 2020 Forum. 2020Orlando, FL: American Institute of Aeronautics and Astronautics, 2020.

[4] Bertrand X, Tost F, Champagneux S. Wing Airfoil Pressure Calibration with Deep Learning[M]//AIAA Aviation 2019 Forum. 2019American Institute of Aeronautics and Astronautics, 2019.

[5] Petermann J, Jung Y S, Baeder J, et al. Validation of Higher-Order Interactional Aerodynamics Simulations on Full Helicopter Configurations[J]. Journal of the American Helicopter Society, 2019, 64(4): 1-13.

[6] Ivanell S, Sørensen J N, Mikkelsen R, et al. Numerical analysis of the tip and root vortex position in the wake of a wind turbine[J]. Journal of Physics: Conference Series, 2007, 75: 012035.

[7] Brunton S, Noack B, Koumoutsakos P. Machine Learning for Fluid Mechanics[J]. Annual Review of Fluid Mechanics, 2020, 52(1): 477-508.

[8] Thuerey N, Weissenow K, Prantl L, et al. Deep learning methods for reynolds-averaged navier-stokes simulations of airfoil flows[J]. AIAA Journal, 2020, 58(1): 25-36.

[9] Bhatnagar S, Afshar Y, Pan S, et al. Prediction of aerodynamic flow fields using convolutional neural networks[J]. Computational Mechanics, 2019, 64(2): 525-545.

[10] Raissi M, Yazdani A, Karniadakis G E. Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations[J]. Science, 2020, 367(6481): 1026-1030.

[11] Duru C, Alemdar H, Baran Ö U. CNNFOIL: Convolutional encoder decoder modeling for pressure fields around airfoils[J]. Neural Computing and Applications, 2021, 33(12): 6835-6849.

[12] Guo X, Li W, Iorio F. Convolutional neural networks for steady flow approximation[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2016: 481-490San Francisco California USA: ACM, 2016: 481-490.

[13] Sekar V, Zhang M, Shu C, et al. Inverse design of airfoil using a deep convolutional neural network[J]. AIAA Journal, 2019, 57(3): 993-1003.

[14] Chen J, Viquerat J, Hachem E. U-net architectures for fast prediction of incompressible laminar flows[Z]. arXiv, 2019(2019-10-25).

[15] Kim M I, Yoon H S. Geometric modification for the enhancement of an airfoil performance using deep CNN[J]. Ocean Engineering, 2022, 266: 113000.

[16] Saetta E, Tognaccini R, Iaccarino G. AbbottAE: An autoencoder for airfoil aerodynamics[C]//AIAA AVIATION 2023 Forum. 2023San Diego, CA and Online: American Institute of Aeronautics and Astronautics, 2023.

[17] Immordino G, Da Ronch A, Righi M. Deep–learning framework for aircraft aerodynamics prediction[C]//AIAA AVIATION 2023 Forum. 2023San Diego, CA and Online: American Institute of Aeronautics and Astronautics, 2023.

[18] Zhou H, Xie F, Ji T, et al. Fast transonic flow prediction enables efficient aerodynamic design[J]. Physics of Fluids, 2023, 35(2): 026109.

[19] Peng X, Kou J, Zhang W. Multi-fidelity nonlinear unsteady aerodynamic modeling and uncertainty estimation based on hierarchical kriging[J]. Applied Mathematical Modelling, 2023, 122: 1-21.

[20] Gennaretti M, Muro D. Multiblade Reduced-Order Aerodynamics for State-Space Aeroelastic Modeling of Rotors[J]. Journal of Aircraft, 2012, 49(2): 495-502.

[21] Wang L, Hu X, Liu X, et al. Blade-vortex interaction detection and extraction under deep neural network-based scale feature model[J]. The Journal of the Acoustical Society of America, 2021, 150(2): 1479-1495.

[22] Chen L-W, Thuerey N. Towards high-accuracy deep learning inference of compressible flows over aerofoils[J]. Computers & Fluids, 2022: 105707.

[23] Pant P, Doshi R, Bahl P, et al. Deep Learning for Reduced Order Modelling and Efficient Temporal Evolution of Fluid Simulations[J]. Physics of Fluids, 2021, 33(10): 107101.

[24] Iungo G V, Santoni-Ortiz C, Abkar M, et al. Data-driven reduced order model for prediction of wind turbine wakes[J]. Journal of Physics: Conference Series, 2015, 625: 012009.

[25] Li R, Zhang Y, Chen H. Transfer learning from two-dimensional supercritical airfoils to three-dimensional transonic swept wings[J]. Chinese Journal of Aeronautics, 2023.

[26] Zahn R, Weiner A, Breitsamter C. Prediction of wing buffet pressure loads using a convolutional and recurrent neural network framework[J]. CEAS Aeronautical Journal, 2023.

[27] Wang X, Zou S, Jiang Y, et al. Swin-FlowNet: Flow field oriented optimization aided by a CNN and swin-transformer based model[J]. Journal of Computational Science, 2023, 72: 102121.

[28] 李润泽;张宇飞;陈海昕. 针对超临界翼型气动修型策略的强化学习[J]. 航空学报, 2021, 42(04): 275-288.

[29] 吕小龙;黄丹;姜冬菊. 基于POD-RBF代理模型的迭代更新反演方法[J]. 计算力学学报, 2022, 39(04): 506-511.

[30] 朱星谕;梅立泉. 基于复合神经网络的多源气动数据建模[J]. 西北工业大学学报, 2024, 42(02): 328-334.

[31] 贾续毅;龚春林;李春娜. 基于POD和BPNN的流场快速计算方法[J]. 西北工业大学学报, 2021, 39(06): 1212-1221.

[32] 周浩. 基于POD降阶的气动软体驱动器快速分析方法研究[D]//西安理工大学. 2023(03)西安理工大学, 2023.

[33] 陈海;钱炜祺;何磊. 基于深度学习的翼型气动系数预测[J]. 空气动力学学报, 2018, 36(02): 294-299.

[34] 金晓威. 物理启发的钝体绕流场机器学习计算方法[D]//哈尔滨工业大学. 2020(02)哈尔滨工业大学, 2020.

[35] 杨华;陈树生;高正红;姜权峰;张伟. 基于贝叶斯框架的旋翼气动力数据融合[J]. 航空学报, 2024, 45(08): 138-150.

[36] 罗杰;段焰辉;蔡晋生. 基于本征正交分解的流场快速预测方法研究[J]. 航空工程进展, 2014, 5(03): 350-357.

[37] Su D, Xu G, Huang S, et al. Numerical investigation of rotor loads of a shipborne coaxial-rotor helicopter during a vertical landing based on moving overset mesh method[J]. Engineering Applications of Computational Fluid Mechanics, 2019, 13(1): 309-326.

[38] Shukla S, Singh S N, Sinha S S, et al. Towards improved understanding of aerodynamic impact of helicopter on ship deck flow environment using SDI model[J]. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 2023, 237(8): 1943-1964.

[39] Krishnan V, Gaonkar G, Motta F C. Optimizing ship airwake database for predicting autospectra using deep learning[J]. Journal of the American Helicopter Society, 2023.

[40] Wang X Q., Sarhaddi D, Wang Z, et al. Novel Reduced Order Modeling Methods for CFD Data: Application to Ship Airwake Data[M]//AIAA Scitech 2019 Forum. 2019American Institute of Aeronautics and Astronautics, 2019.

[41] Xu J, Duraisamy K. Multi-level convolutional autoencoder networks for parametric prediction of spatio-temporal dynamics[J]. Computer Methods in Applied Mechanics and Engineering, 2020, 372: 113379.

[42] Tinney C E, Shipman J, Panickar P. Proper-orthogonal-decomposition-based reduced-order models for characterizing ship airwake interactions[J]. AIAA Journal, 2020, 58(2): 633-646.

[43] Tinney C E, Shipman J, Panickar P. Reduced-order models for characterizing ship airwake interactions.[J]. 2018, 33: 17. .

[44] Sun F, Xie G, Song J, et al. Proper orthogonal decomposition and physical field reconstruction with artificial neural networks (ANN) for supercritical flow problems[J]. Engineering Analysis with Boundary Elements, 2022, 140: 282-299.

[45] Zhu N, Zhang Z, Gnanamanickam E, et al. Space-time characterization of ship airwakes[J]. AIAA Journal, 2023, 61(2): 681-697.

[46] Taymourtash N, Zanotti A, Gibertini G, et al. Stochastic simulation of ship airwake in helicopter shipboard operation[J]. 2021. .

[47] Misaka T. Space-time adaptive model order reduction utilizing local low-dimensionality of flow field[J]. Journal of Computational Physics, 2023, 493: 112475.

[48] Mateer R. The aerodynamics of a modern warship[D]. United Kingdom: The University of Liverpool (United Kingdom), 2020.

[49] Conti P, Guo M, Manzoni A, et al. Multi-fidelity reduced-order surrogate modeling[M]. .

[50] Dooley G M, Krebill A F, Martin J E, et al. Structure of a Ship Airwake at Multiple Scales[J]. AIAA Journal, 2020, 58(5): 2005-2013.

[51] Rajmohan N, Zhao J, He C, et al. Development of a Reduced Order Model to Study Rotor/Ship Aerodynamic Interaction[C]//AIAA Modeling and Simulation Technologies Conference. 2015Kissimmee, Florida: American Institute of Aeronautics and Astronautics, 2015.

[52] Zhu N, Zhang Z, Gnanamanickam E, et al. Dynamics of large-scale flow structures within ship airwakes[M]. .

[53] Deng R, Wang Y, Song Z, et al. Analysis of the characteristics of the spectral orthogonal decomposed flow fields: Numerical and experimental investigations of the air flow field around a simplified container ship model[J]. Ocean Engineering, 2022, 266: 112979.

[54] Guo Z, Xu L, Zhou G, et al. A non-intrusive reduced-order model for wind farm wake analysis based on SPOD-DNN[J]. Wind Engineering, 2023: 0309524X2311626.

[55] Nidhan S, Schmidt O T, Sarkar S. Analysis of coherence in turbulent stratified wakes using spectral proper orthogonal decomposition[J]. Journal of Fluid Mechanics, 2022, 934: A12.

[56] Shukla S, Singh S, Sinha S, et al. An investigation of ship airwakes by scale adaptive simulation[J]. TransNav the International Journal on Marine Navigation and Safety of Sea Transportation, 2020, 14.

[57] Pache R, Rung T. Data-driven surrogate modeling of aerodynamic forces on the superstructure of container vessels[J]. Engineering Applications of Computational Fluid Mechanics, 2022, 16(1): 746-763.

[58] Zhu N. Analysis of Ship Airwakes Using Modal Decomposition[D]. .

[59] Seth D. Contributions to the understanding of ship airwakes using advanced flow diagnostic techniques[D]. .

[60] Gaonkar G H, Mohan R. Extracting analytical models of ship airwake from a database toward qualitative analysis and real-time simulation[C]//65th American helicopter society annual Forum. 2009.

[61] Rajmohan N, Zhao J, He C, et al. An efficient pod based technique to model rotor/ship airwake interaction[C]//Proceedings of the AHS International 68th Annual Forum. 2012(3): 1964-1982.

[62] Yang Y, Zheng Z. Hierarchical encoder-decoder architecture for carrier airwake prediction using attention in frequency domain[C]//2023 2nd Conference on Fully Actuated System Theory and Applications (CFASTA). 2023: 630-635Qingdao, China: IEEE, 2023: 630-635.

[63] 赵所;李震;侯中喜;张大为. 舰尾流场扰动影响分析及抑制技术研究[J]. 华中科技大学学报(自然科学版), 2021, 49(06): 86-91.

[64] 吉洪蕾;陈仁良;李攀; 耦合POD重构舰面流场的直升机舰面起降数值模拟[J]. 航空学报, 2016, 37(03): 771-779.

[65] 张佳佳. 舰船甲板气流场量化品质评估及POD重构研究[D]//大连海事大学. 2018(01)大连海事大学, 2018.

[66] 陶琛. 基于流场仿真的舰船测风位置优化及误差修正研究[D]//哈尔滨工程大学. 2018(01)哈尔滨工程大学, 2018.

[67] 苏萁;王逸斌;赵宁. 基于机器学习算法的舰面流场预警系统研究[J]. 测控技术, 2020, 39(02): 109-114.

[68] 卢垚;刘欢;吴加学. 基于本征正交分解的湍流去噪[J]. 海洋学报, 2022, 44(09): 132-144.

[69] 蒋会明;潘鸿海;闫寒;袁静. 基于SPOD方法的压气机转子叶顶区域非定常流动分析[J]. 热能动力工程, 2023, 38(08): 34-43.

[70] 李凯迪;孙晓晶. 风力机翼型S809绕流流动特性的POD和DMD对比分析[J]. 空气动力学学报, 2024, 42(03): 55-68.

[71] 肖若冶;于剑;马正宵. 卷积自编码器在非定常可压缩流动降阶模型中的适用性[J]. 北京航空航天大学学报, 2022: 1-16.

[72] 徐雨航. 基于深度学习与POD分解的圆柱绕流流场预测研究[D]//哈尔滨工业大学. 2022哈尔滨工业大学, 2022.

[73] 郭婷;夏超;储世俊;杨志刚. 不同转向架构型对高速列车列车风及非定常尾迹的影响[J]. 空气动力学学报, 2022, 40(02): 94-104.

[74] Fernandez N, Owen I, White M, et al. Helicopter Aerodynamic Loading in the Airwake of a Moving Ship[C]//Proceedings of the Vertical Flight Society 79th Annual Forum. 2023: 1-11West Palm Beach, Florida USA: The Vertical Flight Society, 2023: 1-11.

[75] Gnanamanickam E P, Zhang Z, Seth D, et al. Structure of the ship airwake in a simulated atmospheric boundary layer[C]//AIAA AVIATION 2020 FORUM. 2020VIRTUAL EVENT: American Institute of Aeronautics and Astronautics, 2020.

[76] Lin H-H, Wu S-J, Liu T-L, et al. Construction of the Operating Limits Diagram for a Ship-Based Helicopter Using the Design of Experiments with Computational Intelligence Techniques[J]. International Journal of Aeronautical and Space Sciences, 2021, 22(1): 1-16.

[77] Matayoshi N, Forrest J, Hodge S J, et al. Relationship between pilot workload and turbulence intensity for helicopter operations in harsh environments[J]. 2009, 2: 1595-1603. .

[78] Lee D, Horn J F. Simulation of pilot workload for a helicopter operating in a turbulent ship airwake[J]. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 2005, 219(5): 445-458.

[79] Lee D, Horn J, Sezer-Uzol N, et al. Simulation of pilot control activity during helicopter shipboard operations[C]//AIAA Atmospheric Flight Mechanics Conference and Exhibit. 2003Austin, Texas: American Institute of Aeronautics and Astronautics, 2003.

[80] Shukla S, Sinha S S, Singh S N. Ship-helo coupled airwake aerodynamics: A comprehensive review[J]. Progress in Aerospace Sciences, 2019, 106: 71-107.

[81] Vitale A, Corraro G, Corraro F, et al. Ship air-wake identification from experimental data for automatic deck landing and takeoff[J]. Journal of Aircraft, 2023, 60(1): 221-231.

[82] Zhao D. Trajectory optimization and control for autonomous helicopter shipboard landing[D]. Rensselaer Polytechnic Institute, 2021.

[83] Memon W A, Owen I, White M D. SIMSHOL: A predictive simulation approach to inform helicopter–ship clearance trials[J]. Journal of Aircraft, 2020, 57(5): 854-875.

[84] Rafael C F, Da Silva G A L, Guedes M J M. Ship-helicopter operational limitation envelope definition with CFD results and wind tunnel data[J]. Paper ICAS, 2018, 345.

[85] Hodge S J, Zan S J, Roper D M, et al. Time-accurate ship airwake and unsteady aerodynamic loads modeling for maritime helicopter simulation[J]. Journal of the American Helicopter Society, 2009, 54(2): 22005-2200516.

[86] Schau K A, Gaonkar G, Polsky S. Rotorcraft downwash impact on ship airwake: Statistics, modelling, and simulation[J]. The Aeronautical Journal, 2016, 120(1229): 1025-1048.

[87] Memon W A. The development of high-fidelity modelling & simulation for the helicopter ship dynamic interface[D]. .

[88] Zhao D, Krishnamurthi J, Mishra S, et al. A trajectory generation method for time-optimal helicopter shipboard landing[C]//2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). 2019: 7644-7650IEEE, 2019: 7644-7650.

[89] Crozon C, Steijl R, Barakos G N. Coupled flight dynamics and CFD – demonstration for helicopters in shipborne environment[J]. The Aeronautical Journal, 2018, 122(1247): 42-82.

[90] Thedin R, Murman S, Horn J, et al. Effects of atmospheric turbulence unsteadiness on ship airwakes and helicopter dynamics[J]. Journal of Aircraft, 2020, 57: 1-13.

[91] Yuan W, Wall A, Thornhill E, et al. CFD Aided Ship Design and Helicopter Operation[J]. Journal of Marine Science and Engineering, 2022, 10(9): 1304.

[92] Bornemeier M T. A pragmatic analysis of helicopter response to turbulent air wake in a shipboard flight deck environment[D]. .

[93] Sharma A. Development and Application of a Comprehensive Simulation for Modeling Helicopter Ship Landing[D]. .

[94] Hoencamp A, Pavel M D. Concept of a predictive tool for ship–helicopter operational limitations of various in-service conditions[J]. Journal of the American Helicopter Society, 2012, 57(3): 1-9.

[95] Owen I, White M D, Padfield G D, et al. A virtual engineering approach to the ship-helicopter dynamic interface – a decade of modelling and simulation research at the university of liverpool[J]. The Aeronautical Journal, 2017, 121(1246): 1833-1857.

[96] 苏大成. 直升机/舰船动态界面非定常干扰机理及舰艉流动控制研究[D]//南京航空航天大学. 2020(12)南京航空航天大学, 2020.

[97] 胡嘉讯;宗昆;史勇杰;徐国华;赵靖涛. 风-浪-舰耦合对舰船尾流场及直升机气动的影响[J]. 飞行力学, 2023, 41(02): 47-53.

[98] 蒋真理. 高海况舰载机-舰船耦合干扰流场数值模拟分析[D]//南京航空航天大学. 2020(07)南京航空航天大学, 2020.

[99] 李光印;徐国华;史勇杰;苏大成. 主动射流控制对直升机着舰飞行的影响分析[J]. 哈尔滨工业大学学报, 2021, 53(12): 68-79.

[100] 赵俊;马东林;刘纪福;罗骏. 着舰路径对舰载直升机飞行特性的影响[J]. 直升机技术, 2021(04): 18-24+31.

[101] 黄斌. 直升机/舰船耦合流场的CFD模拟及风限图计算[D]//南京航空航天大学. 2016(03)南京航空航天大学, 2016.

[102] 徐雅楠. 舰载直升机安全着舰路径规划与策略研究[D]//南京航空航天大学. 2020(07)南京航空航天大学, 2020.

[103] 胡涛. 舰艉空气流场及舰—机动态配合仿真技术研究[D]//南京航空航天大学. 2016南京航空航天大学, 2016.

[104] Hinton G, Deng L, Yu D, et al. Deep neural networks for acoustic modeling in speech recognition: The shared views of four research groups[J]. IEEE Signal processing magazine, 2012, 29(6): 82-97.

[105] Shafiq M, Gu Z. Deep residual learning for image recognition: A survey[J]. Applied Sciences, 2022, 12(18): 8972.

[106] Min B, Ross H, Sulem E, et al. Recent advances in natural language processing via large pre-trained language models: A survey[J]. ACM Computing Surveys, 2023, 56(2): 1-40.

[107] Wu H, Liu X, An W, et al. A generative deep learning framework for airfoil flow field prediction with sparse data[J]. Chinese Journal of Aeronautics, 2021.

[108] Gupta R, Jaiman R. A hybrid partitioned deep learning methodology for moving interface and fluid–structure interaction[J]. Computers & Fluids, 2021: 105239.

[109] Deng J, Dong W, Socher R, et al. Imagenet: A large-scale hierarchical image database[C]//2009 IEEE conference on computer vision and pattern Recognition. 2009: 248-255Ieee, 2009: 248-255.

[110] Inubushi M, Goto S. Transfer learning for nonlinear dynamics and its application to fluid turbulence[J]. Physical Review E, 2020, 102(4): 043301.

[111] Taud H, Mas J-F. Multilayer perceptron (MLP)[J]. Geomatic approaches for modeling land change scenarios, 2018: 451-455.

[112] LeCun Y, Bottou L, Bengio Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.

[113] Liu Z, Lin Y, Cao Y, et al. Swin transformer: Hierarchical vision transformer using shifted windows[C]//Proceedings of the IEEE/CVF international conference on computer Vision. 2021: 10012-10022.

[114] Han K, Wang Y, Chen H, et al. A survey on vision transformer[J]. IEEE transactions on pattern analysis and machine intelligence, 2022, 45(1): 87-110.

[115] Humphreys G W, Sui J. Attentional control and the self: The self-attention network (SAN)[J]. Cognitive neuroscience, 2016, 7(1-4): 5-17.

[116] Sorkine O. Laplacian mesh processing[J]. Eurographics (State of the Art Reports), 2005, 4(4): 1.

[117] Spalart P, Allmaras S. A one-equation turbulence model for aerodynamic flows[C]//30th aerospace sciences meeting and Exhibit. 1992: 439.

[118] Lock R. Test cases for numerical methods in two-dimensional transonic flows[M]. North Atlantic Treaty Organization, Advisory Group for Aerospace Research …, 1970.

[119] Hassan A A, Straub F K, Noonan K W. Experimental/Numerical Evaluation of Integral Trailing Edge Flaps for Helicopter Rotor Applications[J]. Journal of the American Helicopter Society, 2005, 50(1): 3-17.

[120] Park J J, Florence P, Straub J, et al. Deepsdf: Learning continuous signed distance functions for shape representation[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern Recognition. 2019: 165-174.

[121] Selig M S. UIUC airfoil data site[M]. Urbana, Ill. : Department of Aeronautical and Astronautical Engineering University of Illinois at Urbana-Champaign, 1996-, 1996.

[122] Shields M D, Zhang J. The generalization of Latin hypercube sampling[J]. Reliability Engineering & System Safety, 2016, 148: 96-108.

[123] Paszke A, Gross S, Massa F, et al. PyTorch: An Imperative Style, High-Performance Deep Learning Library[M]. Wallach H, Larochelle H, Beygelzimer A, et al., eds.//Advances in Neural Information Processing Systems 32. 2019: 8024-8035Curran Associates, Inc., 2019: 8024-8035.

[124] Yue Z, Ye F, Zhang Y, et al. Deep Safe Multi-Task Learning[Z]. arXiv, 2021(2021).

[125] Hu R, Singh A. UniT: Multimodal Multitask Learning With a Unified Transformer[C]. .

[126] Salehi Rizi F, Granitzer M. Multi-task Network Embedding with Adaptive Loss Weighting[C]//2020 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM). 2020: 1-5.

[127] Chen H. Multi-resolution Multi-task Network and Polyp Tracking[M]. Bernal J, Histace A, eds.//Computer-Aided Analysis of Gastrointestinal Videos. 2021: 81-89Cham: Springer International Publishing, 2021: 81-89.

[128] Xiao D, Fang F, Pain C C, et al. Non-intrusive reduced order modeling of multi-phase flow in porous media using the POD-RBF method[J]. J Comput Phys, 2015, 1: 1-25.

[129] Hu Z, Xu G, Shi Y. A New Study on the Gap Effect of an Airfoil with Active Flap Control Based on the Overset Grid Method[J]. International Journal of Aeronautical and Space Sciences, 2021, 22(4): 779-801.

[131] Yoon S, Chang L, Kwak D. LU-SGS implicit algorithm for three-dimensional incompressible Navier-Stokes equations with source term[C]//9th Computational Fluid Dynamics Conference. 1989: 1964.

[132] Hu Z, Xu G, Shi Y. A robust overset assembly method for multiple overlapping bodies[J]. International Journal for Numerical Methods in Fluids, 2021, 93(3): 653-682.

[133] Cross J L, Tu W. Tabulation of data from the tip aerodynamics and acoustics test[R]. .

[134] Berkooz G, Holmes P, Lumley J L. The proper orthogonal decomposition in the analysis of turbulent flows[J]. Annual review of fluid mechanics, 1993, 25(1): 539-575.

[135] Tinney C E, Zhao Y. POD-based reduced-order models in translating coordinates[M]//AIAA SCITECH 2023 Forum. 2023American Institute of Aeronautics and Astronautics, 2023.

[136] Gutmann H-M. A radial basis function method for global optimization[J]. Journal of global optimization, 2001, 19(3): 201-227.

[137] Fu R, Xiao D, Navon I M, et al. A data driven reduced order model of fluid flow by Auto-Encoder and self-attention deep learning methods[J]. arXiv:2109.02126 [physics], 2021.

[138] Mason P J. Large-eddy simulation: A critical review of the technique[J]. Quarterly Journal of the Royal Meteorological Society, 1994, 120(515): 1-26.

[139] Canuto V, Cheng Y. Determination of the Smagorinsky–Lilly constant CS[J]. Physics of Fluids, 1997, 9(5): 1368-1378.

[140] Spalart P R, Deck S, Shur M L, et al. A new version of detached-eddy simulation, resistant to ambiguous grid densities[J]. Theoretical and computational fluid dynamics, 2006, 20: 181-195.

[141] Zhang F, Xu H, Ball N. Numerical simulation of unsteady flow over SFS 2 ship model[C]//47th AIAA Aerospace Sciences Meeting Including The New Horizons Forum and Aerospace Exposition. 2009: 81.

[142] Van Muijden J, Boelens O, Van Der Vorst J, et al. Computational ship airwake determination to support helicopter-ship dynamic interface assessment[C]//21st AIAA Computational Fluid Dynamics Conference. 2013San Diego, CA: American Institute of Aeronautics and Astronautics, 2013.

[143] Forrest J S, Owen I, Padfield G D, et al. Ship-Helicopter Operating Limits Prediction Using Piloted Flight Simulation and Time-Accurate Airwakes[J]. Journal of Aircraft, 2012, 49(4): 1020-1031.

[144] Alpman E, Long L N, Bridges D O, et al. Fully-Coupled Simulations of the Rotorcraft / Ship Dynamic Interface[C]//Annual Forum Proceedings-American Helicopter Society. 2007, 1367: 16.

[145] Bridges D, Horn J, Alpman E, et al. Coupled Flight Dynamics and CFD Analysis of Pilot Workload in Ship Airwakes[C]//AIAA Atmospheric Flight Mechanics Conference and Exhibit. 2007Hilton Head, South Carolina: American Institute of Aeronautics and Astronautics, 2007.

[146] Bailey F. A simplified theoretical method of determining the characteristics of a lifting rotor in forward flight[R]. US Government Printing Office Washington, DC, USA, 1941.

[147] Ballin M G. Validation of a Real-Time Engineering Simulation of the UH-60A Helicopter[J]. NASA-TM-88360, 1987: 204.

[148] 薛树强, 杨元喜. 广义反距离加权空间推估法[J]. 武汉大学学报 (信息科学版), 2013, 38(12): 1435-1439.

[149] Lee R G, Zan S J. Unsteady aerodynamic loading on a helicopter fuselage in a ship airwake[J]. Journal of the American Helicopter Society, 2004, 49(2): 149-159.

[150] Lee R G, Zan S J. Wind tunnel testing of a helicopter fuselage and rotor in a ship airwake[J]. Journal of the American Helicopter Society, 2005, 50(4): 326-337.

[151] Memon W A, Owen I, White M D. Motion Fidelity Requirements for Helicopter-Ship Operations in Maritime Rotorcraft Flight Simulators[J]. Journal of Aircraft, 2019, 56(6): 2189-2209.